Research map / Graph learning / Graph algorithms and clustering
Graph-Cut RANSAC
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add spatial-coherence pairwise energy to the local optimization (ablation) | GC-RANSAC · All evaluated problems | geometric error | Helped |
| Use λ = 0.1 for the spatial-coherence weight (ablation) | GC-RANSAC · All evaluated problems | geometric error | Helped |
| Replace a fixed iteration limit with the confidence-increase criterion for applying local optimization (ablation) | GC-RANSAC · All problems combined | processing time | Helped |
| Replace a fixed iteration limit with the confidence-increase criterion for applying local optimization (ablation) | GC-RANSAC · All problems combined | geometric error | No clear effect |
| Add a RANSAC-like procedure using 7m-size samples within local optimization (ablation) | GC-RANSAC · Experiments across tested problems | accuracy | No clear effect |
| Add a RANSAC-like procedure using 7m-size samples within local optimization (ablation) | GC-RANSAC · Experiments across tested problems | processing time | Hurt |
| Use graph-cut local optimization with spatial coherence rather than distance-only inlier selection | GC-RANSAC · 2D line fitting | not-all-inlier samples leading to correct solution: 6% -> 15% | Helped |
| Use graph-cut local optimization with spatial coherence rather than distance-only inlier selection | GC-RANSAC · Fundamental matrix estimation | not-all-inlier samples leading to correct solution: 29% -> 32% | Helped |
| Use GC-RANSAC instead of the strongest listed baseline | fundamental matrix estimation on kusvod2, 60 FPS · kusvod2 (24 pairs) | geometric error: 4.95 -> 4.65 | Helped |
| Use GC-RANSAC instead of the strongest listed baseline | homography estimation on EVD, 60 FPS · EVD (15 pairs) | geometric error: 1.51 -> 1.53 | Hurt |
| Use GC-RANSAC instead of competing methods | 2D line fitting · Synthetic straight and dashed lines | mean angular error | Helped |
| Use GC-RANSAC instead of competing methods | fundamental matrix estimation · Fundamental matrix datasets | geometric error | Helped |
Closest papers by meaning
- Latent RANSAC (CVPR 2018 · similarity 0.86)
- Learning To Find Good Models in RANSAC (CVPR 2022 · similarity 0.85)
- RANSAC Scoring Done Right (NeurIPS 2026 · similarity 0.85)
- CONSAC: Robust Multi-Model Fitting by Conditional Sample Consensus (CVPR 2020 · similarity 0.84)
- MAGSAC: Marginalizing Sample Consensus (CVPR 2019 · similarity 0.84)
- PARSAC: Accelerating Robust Multi-Model Fitting with Parallel Sample Consensus (AAAI 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.